PHYSICAL REVIEW D 103, 063022 (2021)
INTEGRAL x-ray constraints on sub-GeV dark matter
Marco Cirelli ,1,* Nicolao Fornengo ,2 Bradley J. Kavanagh ,3 and Elena Pinetti 1,2 1Laboratoire de Physique Th´eorique et Hautes Energies (LPTHE), UMR 7589 CNRS & Sorbonne University, 4 Place Jussieu, Paris F-75252, France 2Dipartimento di Fisica, Universit`a di Torino & INFN, Sezione di Torino, via P. Giuria 1, Torino I-10125, Italy 3Instituto de Física de Cantabria, (IFCA, UC-CSIC), Av. de Los Castros s/n, Santander 39005, Spain
(Received 2 September 2020; accepted 4 February 2021; published 17 March 2021)
Light dark matter (DM), defined here as having a mass between 1 MeVand about 1 GeV, is an interesting possibility both theoretically and phenomenologically, at one of the frontiers of current progress in the field of DM searches. Its indirect detection via gamma rays is challenged by the scarcity of experiments in the MeV– GeV region. We look therefore at lower-energy x-ray data from the INTEGRAL telescope, and compare them with the predicted DM flux. We derive bounds which are competitive with existing ones from other techniques. Crucially, we include the contribution from inverse Compton scattering on galactic radiation fields and the cosmic microwave background, which leads to much stronger constraints than in previous studies for DM masses above 20 MeV.
DOI: 10.1103/PhysRevD.103.063022
I. INTRODUCTION NEWS-G [21], DARKSIDE [22],EDELWEISS[23],andin the future LBECA [24]),possibly exploitingthe productionof The possibility that dark matter (DM) consists of a light detectable signals via DM-electron scattering [25–28],the particle has gained increasing attention recently (for def- Migdal effect1 [15,36,37] or the interactions of DM with the initeness we intend here a mass between a few MeV and Sun or with cosmic rays [38–50]. Efforts are underway as well about a GeV, as we discuss below). Searches for DM have to develop new detection strategies [51,52], including the use long been dominated by the paradigm of (heavier) weakly of semiconductors [53–55] (such as in the SENSEI [56] and interacting massive particles (WIMPs) [1,2], but with no DAMIC [57,58] experiments), superconductors [59,60], convincing WIMP signal observed so far in direct detection superfluid helium [61–66], evaporating helium [67],2D [3], indirect detection [4–6] or collider searches [7,8], the materials such as arrays of nanotubes and graphene attention is turning to lighter (or heavier) candidates. Light [68,69], 1D materials such as superconducting nanowires DM is indeed in some sense a new frontier for searches, [70], chemical bond breaking [71], production of color-center requiring new analysis strategies and experimental tech- defects in crystals [72], polar materials (i.e., crystals with niques to achieve sensitivity [9]. easily excitable polar atomic bonds [73,74]), paleodetectors In direct detection (DD), most current experiments lose [75,76], magnetic bubble chambers [77], Casimir forces sensitivity for DM masses below ∼1 GeV. This is because the between nucleons [78], aromatic targets [79],molecular standard method of detection relies on detecting the small gas excitations [80], or diamonds [81]. Such a long list amounts of energy deposited by DM via nuclear recoils, illustrates the strong interest in the DD community for this which becomes ineffective for DM much lighter than a typical regime, and bears well for the future. nucleus. However, many significant efforts are underway to On the side of collider searches, the situation has some explore the sub-GeV regime. This includes extending the resemblance with the DD case. The current experiments sensitivity of “traditional” nuclear recoil detectors to ultralow and flagship colliders (essentially the LHC) are not suitable energy thresholds [10,11] (XENON-10 [12], XENON-1T [13],LUX[14,15], CRESST [16–19], SUPER-CDMS [20], 1In the standard DD method, the ionization signal is typically produced as the nucleus struck by the DM particle recoils and hits *[email protected] in turn other nuclei, freeing their electrons. This requires a certain non-negligible energy deposition. The Migdal effect [29–35] Published by the American Physical Society under the terms of refers instead to the fact that even a slowly recoiling or just the Creative Commons Attribution 4.0 International license. shaken nucleus can lose one or some of its electrons, and Further distribution of this work must maintain attribution to therefore produce a signal. Standard ionization is an in-medium the author(s) and the published article’s title, journal citation, process, while the Migdal effect can take place even for the and DOI. Funded by SCOAP3. scattering of an isolated atom.
2470-0010=2021=103(6)=063022(18) 063022-1 Published by the American Physical Society CIRELLI, FORNENGO, KAVANAGH, and PINETTI PHYS. REV. D 103, 063022 (2021)
2 for exploring the low mass region: the staple signature of value E0 to a final value E ≈ 4γ E0 upon scattering off an DM production is the missing energy corresponding to the electron with relativistic factor γ ¼ Ee=me.Hence,a1GeV DM mass, which in the sub-GeV case is swamped in the electron will produce a ∼1.5 keV x ray when scattering off −4 experimental background. The strategy is therefore to turn the CMB (E0 ≈ 10 eV). By the same token, a mildly to the search for associated states, i.e., particles that belong relativistic MeV electron will produce a ∼0.15 keV x ray “ ” to a new dark sector, more extended than just the DM when scattering off UV starlight (E0 ≈ 10 eV). These con- particle. In particular, if sub-GeV DM is to be produced by siderations roughly define our range of interest for DM ≃ 1 → 1 4 thermal freeze-out in the early Universe, this requires the masses: mDM MeV GeV. So our goal is to explore existence of new light force mediators (“dark photons”). whether there are cases in which x-ray observations can These mediators provide a link between the dark and impose constraints on sub-GeV DM that would otherwise visible sectors and are therefore subject to active searches. fall below the sensitivity of the more conventional gamma- Indeed many theory analyses [82–87] and experimental ray searches. To this end, we focus on data from the projects [88–91] pursue this direction. For a detailed and INTEGRAL x-ray satellite [121]. recent review, see, e.g., Ref. [92], where additional refer- The electrons and positrons produced by the annihilation ences can be found. of light DM propagate in the Galaxy. The dominant energy In indirect detection (ID), one typically searches for the loss processes for these particles at low energy are Standard Model particles (charged particles, such as elec- bremsstrahlung, ionization and Coulomb scattering. trons and positrons, or neutral ones, such as gamma rays These therefore reduce the power that is effectively injected and neutrinos) produced in the annihilation of DM in the into ICS. However these processes are relevant where gas is Galaxy, with energies at or just below the DM mass. present, i.e., essentially in the Galactic center and within the Concerning charged particles, the problem is that solar disk. We will therefore focus on observational windows at activity holds back sub-GeV charged cosmic rays and 2 mid-to-high Galactic latitudes, where indeed ICS (together therefore we have no access to them. Concerning gamma with synchrotron emission) is a relevant mechanism of rays, the sensitivity of the most powerful of the recent photon production. telescopes, Fermi-LAT, stops at about 100 MeV, so that From the theoretical point of view, the case for sub-GeV typically DM masses only as low as about 1 GeV can be DM is arguably not as strong as the traditional case for probed. At much lower energies, below a few MeV, one has WIMPs. On the other hand, considering the lack of evidence ∼1 competitive data from INTEGRAL. But in between and for a DM in the typical WIMP mass range, it is definitely 250 MeV, only relatively old data from COMPTEL are worth leaving no stone unturned [122]. From the phenom- available and no current competitive experiment exists. enological perspective, light (scalar) DM has been invoked as Indeed, a number of authors have discussed proposals to fill a viable possibility [123–125] and as an explanation this so-called MeV gap in a useful way for DM searches [126–129] of the 511 KeV line from the center of the galaxy – [94 106]. Alternatively, low energy neutrinos have been [130]. From the point of view of theoretical motivations, well considered, but they are found not to be very competitive founded models include for instance the strongly inter- with respect to photons (e.g., the projected sensitivity of acting massive particles (SIMP) scenarios [131–135],the 20 years of run with the future Hyper-Kamiokande detector WIMP-less idea [136], forbidden dark matter [137], axinos is weaker than the existing bound that we will discuss [138,139],neutrinomassgeneration[128,140],andeven below, and the new ones we will derive) [107,108]. certain supersymmetry configurations [141–143]: all these Another possibility for ID of such light DM (which is the constructions naturally include or even predict sub-GeV DM one we will entertain here) is to look at a range of energies 3 particles. One may also add to the list asymmetric DM [144]: much lower than that of the DM mass. The basic idea is the in the standard lore it naturally predicts DM with a mass of a following: the electrons and positrons produced in the few GeVand no annihilation signals, but actually the mass can Galactic halo by the annihilations of DM particles with a ≃ 1 ≲ 1 be lighter [145,146] and residual or late annihilations can mass m GeV have naturally an energy E GeV; they occur and give rise to an interesting phenomenology undergo inverse Compton scattering (ICS) on the low energy [145,147]. Freeze-in DM production [148] is also often cited photons of the ambient bath (the cosmic microwave back- in connection to sub-GeV DM, not necessarily because ground (CMB), infrared light and starlight) and produce x freeze-in points to the sub-GeV range but because it provides rays, which can be searched for in x-ray surveys. Indeed, the a viable DM production mechanism. ICS process increases the photon energy from the initial low More broadly, studies addressing the motivations, the phenomenology and the constraints (notably from cosmol- 2An exception to this point is the use of data from the Voyager ogy) of MeV–GeV DM include [149–165]. spacecraft, which is making measurements outside of the helio- sphere [93]. We will comment on the corresponding constraints later. 4Note that we are not interested in keV DM, that can, e.g., 3For former applications of the same idea to heavy, WIMP- produce x rays by direct annihilation or decay. That is a whole like, DM see for instance [109–118]. other set of searches, e.g., for keV sterile neutrino DM [119,120].
063022-2 INTEGRAL X-RAY CONSTRAINTS ON SUB-GeV DARK MATTER PHYS. REV. D 103, 063022 (2021)
The rest of this paper is organized as follows. In Sec. II subdominant outside of the gas-dense region of the Galactic we review the computation of x-rays from DM annihila- Center that we do not consider, as mentioned above. tions, via the ICS process. In Sec. III we present the x-ray The differential flux of the final state radiation or measurements by INTEGRAL that we employ. In Sec. IV radiative decays photons from the annihilations of a DM we present the results, in particular deriving constraints in particle of mass mDM is computed via the standard the usual plane of annihilation cross section versus mass of expression (see, e.g., Ref. [169]): the DM particle, and we briefly compare with other existing 2 f constraints. In Sec. V we summarize and conclude. dΦ γ 1 ρ dN γ FSR r⊙ ⊙ θ σ FSR Ω ¼ 2 4π Jð Þh vif ; dEγd mDM dEγ II. X RAYS FROM DM ANNIHILATIONS Z ds ρðrðs; θÞÞ 2 JðθÞ¼ ; ð4Þ In this section we briefly review the basic formalism for l:o:s: r⊙ ρ⊙ (soft γ ray and) hard x-ray production from DM annihi- 3 lations, essentially via final state radiation and ICS emis- where r⊙ ≃ 8.33 kpc and ρ⊙ ¼ 0.3 GeV=cm are the sion. For all details we refer the reader to Refs. [166,167], conventional values for the distance of the Sun from the whose notation we mostly follow. Two important dif- Galactic Center (GC) and the local DM density. Here, Eγ ferences however apply. First, Refs. [166,167] do not deal denotes the photon energy, dΩ is the solid angle, r is the with DM masses smaller than 5 GeV, so that we cannot distance from the GC in spherical coordinates, in turn straightforwardly use the tools developed there. Second, in expressed in terms of s, the coordinate that runs along the Refs. [166,167] a detailed treatment of the energy-loss and line of sight (s ¼ 0 corresponds to the Earth position), and diffusion process is employed. Since, as mentioned above, 2 2 2 the angle θ: r ¼ s þ r⊙ − 2sr⊙ cos θ. We use natural we are mostly interested in higher latitude signals where units c ¼ ℏ ¼ 1 throughout the paper. The normalized J ICS energy losses dominate, we adopt a simplified treat- factor corresponds to the integration along the line of sight ment (described below). of the square of the galactic DM density profile ρðrÞ, for Let us consider a direction of observation from Earth that which we adopt a standard Navarro-Frenk-White profile θ is identified by the angle (the aperture between the [166,170] direction of observation and the axis connecting the Earth to the Galactic Center), or, equivalently, by the latitude and −2 l ρ ρ rs 1 r longitude pair ðb; Þ. At each point along this direction, NFWðrÞ¼ s þ with DM particles annihilate, contributing to the photon signal, r rs ρ 0 184 3 24 42 which we collect by integrating all the contributions along s ¼ . GeV=cm ;rs ¼ . kpc: ð5Þ the line of sight. Since we focus on DM lighter than a few GeV, we consider only three annihilation channels: The annihilation cross section in each of the final states f of – σ Eqs. (1) (3) is denoted h vif. → þ − DM DM e e ; ð1Þ For the spectrum of FSR photons we adopt the following expressions, which we derive from the analysis of [171] þ − DM DM → μ μ ; ð2Þ and reproduce here for the convenience of the reader:
− → πþπ − DM DM ; ð3Þ dNlþl α 1 ν FSRγ A þ Rð Þ− 2B ν ¼ 2 ln Rð Þ ; ð6Þ dEγ πβð3 − β Þm 1 − RðνÞ which are kinematically open whenever mDM >mi (with DM i ¼ e, μ, π) and that we study one at a time. The pion channel is representative of a hadronic DM channel. We do where not consider the annihilation into a pair of neutral pions, since in this case the (boosted to the DM frame) γ rays do 1 β2 3 − β2 A ð þ Þð Þ − 2 3 − β2 2ν not reach down to the energies covered by INTEGRAL. ¼ ν ð Þþ ; ð7Þ For each channel, the total photon flux is given by the sum of two contributions: the emission from the charged 3 − β2 particles in the final state (from final state radiation, FSR, B ¼ ð1 − νÞþν ; ð8Þ and, whenever relevant, other radiative decays, Rad) and ν the photons produced via ICS by DM-produced energetic μ ν β2 e . Also, another contribution to the total photon flux is where l ¼ e, and we have defined: p¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiEγ=mDM, ¼ þ 1–4μ2 μ 2 ν 1 − 4μ2 1 − ν given by the in-flight annihilation of DM-produced e with with ¼ ml=ð mDMÞ and Rð Þ¼ =ð Þ. ambient e− from the gas [168]. This process is however For the pion [171]
063022-3 CIRELLI, FORNENGO, KAVANAGH, and PINETTI PHYS. REV. D 103, 063022 (2021)
− dNπþπ 2α ν 1 − ν 1 β2 1 ν FSRγ − ν þ − 1 þ Rð Þ ¼ πβ 2 ν Rð Þþ 2ν ln 1 − ν ; ð9Þ dEγ mDM β Rð Þ with the same definitions as above with ml → mπ. − − The muon can undergo the radiative decay μ → e ν¯eνμγ, and this channel can be a relevant source of soft photons. For the photon spectrum we adopt the parametrization of Refs. [172,173] which in turn are derived from Ref. [174]. In the muon rest frame the expression we adopt is therefore μ α 1 − 1 − dNRad γ ð xÞ 2 x ¼ 12ð3 − 2xð1 − xÞ Þ log þ xð1 − xÞð46 − 55xÞ − 102 ; ð10Þ dEγ 36πEγ r Eμ¼mμ
2 max where x ¼ 2Eγ=mμ, r ¼ðme=mμÞ and the maximal photon energy is Eγ ¼ mμð1 − rÞ=2 ≃ 52.8 MeV. For muons in → μþμ− flight, Eq. (10) is boosted to the frame where the muon has energy Eμ ¼ mDM. For the process DM DM ,a multiplicity factor of 2 needs to be applied, since a pair of muons is produced for each annihilation event. − − Charged pions can also produce photons radiatively through the process π → l ν¯lγ with l ¼ e, μ. Also in this case we adopt the parametrization of Ref. [172], which has been derived from Ref. [175]. In the pion rest frame the expression we adopt is then
π α dNRad γ ½fðxÞþgðxÞ ¼ 2 2 2 ; ð11Þ dEγ 24πmπfπ r − 1 x − 1 rx Eπ ¼mπ ð Þ ð Þ
2 where x ¼ 2Eγ=mπ, r ¼ðml=mπÞ , fπ ¼ 92.2 MeV is the pion decay constant and
2 4 2 2 2 2 fðxÞ¼ðr þ x − 1Þ½mπx ðF þ F Þðr − rx þ r − 2ðx − 1Þ Þ ffiffiffi A V p 2 − 12 2fπmπrðx − 1Þx ðFAðr − 2x þ 1ÞþxFVÞ 2 2 −24fπrðx − 1Þð4rðx − 1Þþðx − 2Þ Þ ; ffiffiffi p 2 r 2 gðxÞ¼12 2fπrðx − 1Þ log ½mπx ðF ðx − 2rÞ − xF Þ 1 − x A V ffiffiffi p 2 2 þ 2fπð2r − 2rx − x þ 2x − 2Þ ; ð12Þ
2 2 where the axial and vector form factors are FA ¼ 0.0119 and FVðq Þ¼FVð0Þð1 þ aq Þ with FVð0Þ¼0.0254, a ¼ 0.10, q2 ¼ð1 − xÞ [172,176]. When the pion decays into an on shell muon, the radiative decay of the muon is again a source of low energy photons. Following Ref. [172], the total radiative charged pion spectrum in the pion rest frame can be phrased as
π π μ dN X dN dN Rad tot γ π → lν Rad γ π → μν Rad γ ¼ BRð lÞ þ BRð μÞ ; ð13Þ dEγ l μ dEγ dEγ Eπ ¼mπ ¼e; Eπ¼mπ Eμ¼E⋆
2 2 where E⋆ ¼ðmπ þ mμÞ=ð2mπÞ is the muon energy in the in the derivation on the DM annihilation cross section pion rest frame. For the process DM DM → πþπ−, where bounds in Sec. IV. the pion is in flight, Eq. (13) is boosted to the frame where The differential flux of the inverse Compton scattering Eπ ¼ mDM and a multiplicity factor of 2 is applied, since photons received at Earth is written in terms of the also in this case a pair of pions is produced for each emissivities jðEγ; x⃗Þ of all cells located along the line of annihilation event. As pointed out in Ref. [172], the muon sight at position x⃗: Z radiative decay provides a quite relevant soft-photon Φ ⃗ l d ICγ 1 jðEγ; xðs; b; ÞÞ production channel. Both in the muon and in the pion ¼ ds : ð14Þ dEγdΩ Eγ 4π annihilation channel, the Rad photon flux arising from the l:o:s: muon radiative decay is dominant over the FSR emission Here the spherical symmetry of the system around the for DM masses up to a few GeV. This will be explicitly seen center of the Galaxy is broken by the distribution of the
063022-4 INTEGRAL X-RAY CONSTRAINTS ON SUB-GeV DARK MATTER PHYS. REV. D 103, 063022 (2021)
10–13 In the Gal Plane: Above the Gal Center: 10–14 (R,z) = (8,0) kpc (R,z) = (0,4) kpc
10–15 ] 10–16
–17
[GeV/sec 10 input e± energy: (R,z)=(8.33,0) kpc 10–11 b 100 MeV 1 GeV 10 GeV 10–18 10–12 CMB 10–19 10 MeV –13 10 10–20 IR –21 10–14 10 Ioniz Ioniz
1 MeV Energy loss coefficient
ICS power [1/sec] –22 Brems Brems –15 10 10 Opt ICS ICS –23 10 Synchrotron Synchrotron 10–16 –24 –6 –5 –4 –3 –2 –1 2 3 4 10 10 10 10 10 10 10 1101010 10 1 10 102 103 104 1 10 102 103 104 photon energy E [MeV] e± energy [MeV] e± energy [MeV]
FIG. 1. Left: ICS power as a function of the emitted photon energy, for the listed input e energies. In the 10 GeV case, for illustration we show the contributions of the three components of the photon bath: CMB, infrared light and optical light. Right: energy losses for electrons and positrons in the energy regime of interest. The first subpanel refers to the situation in the Galactic plane, in a spot similar to the location of the solar system: one sees that the e lose most of their energy via interactions with the gas (ionization and bremsstrahlung). The second subpanel is for a point well outside the plane, on the vertical of the Galactic Center: one sees that the ICS emission is dominant down to ∼40 MeV. Lines of sight that avoid the Galactic plane (i.e., at high latitude) are therefore preferred for our purposes. ambient light, which mostly lies in the galactic disk. Hence and references therein) but, for reference, we plot the total we use the 3D notation x⃗instead of r, and we need the power P in the energy regime of interest in Fig. 1, left. In l θ σ ðb; Þ coordinates instead of the single angle . These the above equation, IC is the Klein-Nishina cross section. quantities are related by the usual expression cos θ ¼ The local e spectral number density we are concerned cos b cos l. It will also be convenient to work in cylindrical with here is the result of injection due to DM annihilation, coordinates ðR; zÞ. The emissivity is obtained as a con- plus the subsequent diffusion of the e in the local Galactic volution of the density of the emitting medium with the environment, subject to energy losses by several processes power that it radiates. In this case therefore (see below). We adopt a simplified treatment by neglecting Z diffusion, i.e., we assume that electrons and positrons m DM dne scatter off the ambient photons in the same location where jðEγ; x⃗Þ¼2 dE P ðEγ;E ; x⃗Þ ðE ; x⃗Þ; ð15Þ e IC e e me dEe these e were produced by annihilation events of DM particles. This is sometimes referred to as the “on the spot” where Ee denotes the electron total energy and dne =dEe is approximation. While this is certainly a drastic approxi- the local electron (or positron) spectral number density (the mation, we verified a posteriori that it is justified in the overall factor of 2 takes into accountP the equal populations regimes of our interest: the timescales of diffusion over P Pi of electrons and positrons). ¼ i IC is the differential lengths comparable to the angular bins of the INTEGRAL power emitted into photons due to ICS radiative processes data are longer than the loss timescales, except for lines of Z sight towards high latitudes, which however contribute Pi ⃗ ϵ ϵ ⃗ σ ϵ little to the signal. In this regime, the quantity can be simply ðEγ;Ee; xÞ¼Eγ d nið ; xÞ ICð ;Eγ;EeÞ; ð16Þ IC expressed as (see, e.g., [115]) Z where the sum on i runs over the different components of dn 1 mDM e ðE ; x⃗Þ¼ dE˜ Q ðE˜ ; x⃗Þ; the photon bath, expressed by their photon number den- dE e b ðE ; x⃗Þ e e e sities n (ϵ is the photon energy before the scattering): e tot e Ee i σ ρ ⃗ 2 CMB, dust-rescattered infrared light (IR) and optical star- ˜ ⃗ h vi ðxÞ dNe with QeðEe; xÞ¼ 2 ˜ : ð17Þ light (SL). For the last two components, we use the mDM dEe interstellar radiation field (ISRF) maps extracted from ⃗ ≡ − dE [177],asin[115,166,167]. We do not detail here further Here btotðE; xÞ dt ¼ bCoulþioniz þ bbrem þ bsyn þ bICS the computations (we refer the reader to [115,166,167,178] is the energy loss function, which takes into account all
063022-5 CIRELLI, FORNENGO, KAVANAGH, and PINETTI PHYS. REV. D 103, 063022 (2021)
FIG. 2. Example photon spectra from sub-GeV DM that illustrate our main points. Left: the hard x-ray and soft γ-ray spectrum produced by a 150 MeV DM particle annihilating into μþμ−. We show the different components in color and the total flux in thick black. The spectrum cuts off before reaching Fermi’s data (taken from [182] and reported here just for reference, as they are not the focus of our work), but produces a signal in x rays that can be constrained by INTEGRAL (taken from [121], Fig. 7). The FSR (blue dashed) and Rad (blue dot- dashed) contributions yield signals that pass well below the x-ray data. However, the inclusion of the DM-induced ICS contribution on the different components of the Galactic ambient light (StarLight (SL, green dotted), dust-reprocessed infrared light (IR, brown dotted) ) and the CMB (not visible in the plot), leads to a flux which is orders of magnitude larger, thus producing stronger constraints. Right: the same for a 10 MeV DM particle annihilating into eþe−. In this case the limit is instead driven by the FSR contribution because the DM ICS contributions fall to too low energy for INTEGRAL. In these illustrations, the signals are computed over the jbj < 15°; jlj < 30° region of interest: in our analysis we actually use smaller regions of interest, removing low latitudes (see Sec. III).